James E. Campbell* The Republican Wave of 2014: The Continuity of the 2012 and 2014 Elections

Transcription

1 The Forum 2014; 12(4): James E. Campbell* The Republican Wave of 2014: The Continuity of the 2012 and 2014 Elections Abstract: This article examines the influences on the 2014 midterm congressional elections. The election was a decisive victory for the Republican Party. Republicans picked up nine Senate seats and with them control of the Senate. They also gained thirteen House seats, extending their House majority to its largest size in over 80 years. Although Republican seat gains can be explained at one level by the normal processes of surge and decline plus the midterm as a referendum on the performance of the presidential party, this does not explain why the midterm referendum was so lopsidedly against the Democrats that it created a Republican wave in the election. This analysis traces the reasons for the Democrats shellacking to the 2012 presidential election. The reason for President Obama s re election in 2012 has been misinterpreted as an ideological triumph or as a positive referendum on the president s job performance. Neither was the case. As a first party-term incumbent, President Obama was given additional time to revive the weak economy that he inherited from President Bush. The 2012 referendum on the president s performance was a suspended referendum, not a vote of confidence. The analysis shows that the economy was weak by historical standards in 2012 and remained so in 2014 and that voters realized that the economy was weak in 2012 and continued to believe so in The crucial political difference was that voters blamed former President Bush in 2012, but held the Democrats responsible in DOI /for Introduction The 2014 midterm elections were a decisive victory for the Republican Party. The focal point of the election was the hotly contested battle for control of the Senate. Republicans picked up nine Senate seats, three more than needed to shift control. Of 36 seats up for election, Democrats had held 21 and Republicans 15. *Corresponding author: James E. Campbell, Department of Political Science, 520 Park Hall, University at Buffalo, SUNY, Buffalo, NY , USA, jcampbel@buffalo.edu

2 610 James E. Campbell Voters more than evened things up. Coming out of the election, Republicans won 24 seats and Democrats held on to only 12 (all but one an incumbent). The new Senate is Republican, counting as Democrats the two independents who caucus with the Democrats. In the House, Republicans extended their majority to a historic level. With the addition of 13 members to the Republican side of the aisle, the GOP majority now stands at 247 members. It is the largest Republican House majority in generations. It is one larger than that elected in The last GOP majority that was larger was elected in 1928 when Herbert Hoover defeated Al Smith in a landslide, prohibition was the law of the land, and talking motion pictures were in their infancy. Republicans even gained governorships, despite having more to defend. In short, Democrats were delivered an across-the-board thumping. It was a Republican wave. A Midterm of Change, But What Changed? The results of the 2014 midterm election are clear. They are a dramatic shift from the 2012 presidential election. In 2012, voters reelected President Obama, trimmed the Republicans House majority by eight seats and added two Democrats to the Senate. Two years later, the president s approval ratings had dropped into the low to mid 40-percent range and voters handed Democrats a drubbing. It was a huge turnaround. But while the results of the 2014 midterm are clear, the reasons for them are not. They present something of a puzzle. Prior to the 2014 election, public sentiment was with the Republicans. Yet without a crisis or some other dramatic development, the inspiration for a Republican wave seemed lacking (Abramowitz 2014; Campbell 2014a). Observing President Obama s short coattails in 2012 and the Democrats already low numbers in the House, Alan Abramowitz concluded that Republican House gains would be minimal and that there was no wave in sight (Abramowitz 2014, p. 772). Along with others, I noted that there were few split-result districts for Republicans to pick off and that there appeared to be no disruptive wave on the horizon (Campbell 2014a). Even so, several election forecast models, including my own, detected the Republican wave far in advance of the election. My seats in trouble forecasts missed the House outcome by only three seats (a forecast of +16 for the Republicans) and the Senate outcome by only a single seat (+8 Republicans) (Campbell 2014b). But what the forecasts did not reveal was why voters were throwing the Democrats out. What changed between 2012 and 2014 to explain such different electoral verdicts? Did the demographics of the electorate change in ways favoring Democrats

3 The Republican Wave of in 2012 and Republicans 2 years later? Did voters grow more conservative between the elections? And if there were composition changes, why did they change? Were there changing circumstances, policy initiatives, or events that soured voters on the Democrats or drew voters to the Republicans? How did American politics move from Democrats rejoicing in 2012 to Republicans celebrating in 2014? Like everything else, elections can be explained at varying depths. At the surface level, 2014 was another normal midterm election. The standard processes of surge and decline of the political climate from the prior presidential election combined with the midterm referendum on the presidential party s performance to produce the election s outcomes. Nothing out of the ordinary here. The driving force in the 2014 outcomes was the midterm referendum on the record of the presidential party. The next section of the analysis will review these factors in more detail. The second level of explaining why voters rejected the Democrats in the 2014 midterm raises the more interesting question. This is the more intriguing question because not much changed between the 2012 presidential election and the 2014 midterm election. A lack of change in conditions and the electorate somehow begat a great deal of change in the election s outcome. As the analysis will contend, the key to understanding the 2014 midterm election is in a proper understanding of what happened and what did not happen in the unusual presidential election of Presidents are usually elected because voters like the president s values (ideological agreement) or because voters think that the president s party has performed well on matters all Americans care about (valence issues, notably peace and prosperity). What was unusual about President Obama s reelection in 2012 was that voters neither particularly agreed with his values nor did they think that he and his party had governed very effectively, at least not effectively by normal standards applied to previous presidents (Campbell 2013a,b). As my post-2012 election analysis found and as this analysis will again show, President Obama was narrowly reelected because voters are strongly inclined to give a first party-term incumbent more time to get an inherited mess straightened out. President Obama was elected initially in 2008 in the midst of the Wall Street financial system meltdown and the Great Recession. Four years later, a large majority of voters were still not happy with the way things were going. They were particularly dissatisfied with the economy in 2012, but they were willing to give President Obama and his party more time. After another 2 years, the economy was still not appreciably stronger. Time was up. Not much had changed. This lack of change was the problem, the economy remained weak, and this was the basis for the Republican wave of 2014.

4 612 James E. Campbell A Normal Midterm Election At its most basic level, 2014 was a normal midterm election, subject to the normal two forces structuring electoral change in midterms. The outcomes of midterm elections are determined by the prevailing political climate in the prior election for the office and by the current political climate in the midterm. It is that simple. Change is a consequence of the where you start from and where you end up. This is as true about examining the influences on seat change as it is about calculating the extent of seat change. Since prior presidential elections (immediately prior in House midterm elections and 6 years prior in Senate midterms) are more intense events with a stronger flow of information and are almost always in favor of the winning presidential party in House elections, they account for the president s party quite consistently losing House seats in the following midterm elections. Down-ticket candidates benefitting initially from a good year for their party and its standard-bearer are often in trouble when they next must run without such favorable conditions. This was observed in presidential coattail theories of midterm losses (Bean 1950) and in more sophisticated theories of surge and decline (Campbell 1960, 1987, 1997). 1 The presidential surge is often responsible for the arithmetic of seat change, as well. That is, the prior presidential surge adds seats to the in-party that the party must defend in the next election. More seats to defend means subsequently greater potential losses and smaller potential gains. A party cannot gain seats it already holds and a party holding fewer seats has more room to grow (Oppenheimer, Stimson, and Waterman 1986; Campbell 1997). For the 2014 midterm elections, the national political conditions in their prior election years suggested that congressional Democrats would likely sustain losses and that their problems would be proportionately larger in the Senate than in the House. Since President Obama was reelected by a relatively narrow margin in 2012 (with 52% of the two-party popular vote), the election pertinent to House seat change, there had been less of a surge from which to decline fewer seat changes in the House could be expected. Already near their peak of House members in recent decades, it would be difficult for Republicans to do better than to gain back much more than the eight seats they had lost to Democrats in Surge and decline applies (on a 6-year cycle) to Senate elections, but less well than it does in the House since not all states have Senate seats up for election in any year and the national presidential vote indicative of the political climate in an election may be imperfectly reflected in Senate elections because of their representation of states regardless of their populations.

5 The Republican Wave of Prospects looked bleaker for Democrats in the Senate. The class of Senators up for election in 2014 had last been elected in the exceptional presidential election of Early signs that year had suggested a nearly balanced political climate, but that all changed with the unprecedented financial collapse in the midst of the fall campaign (Campbell 2008). The post-collapse climate of 2008 was decidedly in the Democrats favor. This yielded a big class of Democratic Senators seeking reelection in 2014, now without the pro-democratic surge that had helped them in The second force shaping midterm election change is the political climate of the midterm election year itself. Midterm elections are to a significant extent referenda on the performance of the president and his party (Tufte 1975; Campbell 1985). Whether voters approve of the performance of the president and his party because they provided peace and prosperity or disapprove of their performance because they did not, or if they have advanced immoderate policies or for any other reason, midterm elections offer voters an opportunity to make a mid-course correction. 2 The electorate s evaluation of the president s job performance is generally measured by presidential approval ratings in national surveys. Though these approval ratings are reasonable measures of the political climate for the inparty at the midterm, they are complicated by the lack of a politically meaningful neutral point. Since elections are determined by simple majorities of two-party votes (with the electoral vote complication in presidential elections and a few majority rule and runoff systems duly noted), the 50-percent standard for approval ratings is often naively used. But when the president is not up for election and faces no opponent, the 50-percent mark in a survey is arbitrary and meaningless. The approval of only half of the public at the midterm is not politically neutral. No presidential party does well with a president having only a 50-percent approval rating at the midterm. It is relatively cost-free for those who might otherwise oppose the president if he were up for election to blithely say they approve of 2 Some have specified policy-balancing as a separate force in explaining midterm electoral change. It is difficult to see, however, why dissatisfaction with the presidential party s ideological position (too conservative or too liberal) is any different from dissatisfaction of any other sort with the presidential party. If many voters were upset with the president s party because of its immoderate policies or proposed policies, this is easily expressed in lower presidential approval measures. There seems to be no theoretically justifiable inclusion of policy-balancing apart from the referendum-approval variable. Moreover, there is no reason why policy-balancing effects should be a constant, as they are commonly specified. The conditions eliciting policy-balancing voting should vary with the ideological proclivities of presidents and the ideological distributions of both the public and congressional representatives.

6 614 James E. Campbell his job performance. Outside of the conflict of a presidential race, Americans are generally inclined to look positively on their president. 3 Using history as a guide, the politically neutral presidential approval rating in midterms is in the mid- 60-percent range. When the presidential approval ratings have been in this range, the in-party often sustained only minor seat losses (e.g., Kennedy in 1962, Reagan in 1986) or on rare occasions actually gained a few seats (e.g., Clinton in 1998, G.W. Bush in 2002). President Obama s approval rating in the 2014 midterm was 43% in the Gallup Poll, 44% in the National Exit Polls. 4 A presidential approval rating twenty points or so below the politically neutral point indicates that the political climate in 2014 was bleak for Democrats, favorable for Republicans. The mix of surge and decline (stronger in the Senate than in the House) and the political climate of 2014 accounts for Democrats sustaining moderate-sized losses in the House (though historic because of their level) and substantial losses in the Senate. What remains to be answered, however, is why the public rejected President Obama and his party? What changed between the reelection of President Obama in 2012 and the Democrats midterm debacle of 2014? To explain the 2014 midterm at this deeper level, we must return to the often misunderstood presidential election of The Suspended Referendum of 2012 Elections are complex enough and politically contentious enough that they are routinely interpreted in different and often wishful ways. This was certainly the case in Some claimed that the election turned on values, that it was an endorsement of progressive politics. E.J. Dionne of The Washington Post, for instance, wrote that it was a victory for the younger, highly diverse, and broadly progressive America, that the president took on an increasingly militant conservatism intent on vastly reducing the responsibilities of government and cutting taxes even more on the wealthiest Americans (Dionne 2012). He went on 3 Nowhere is this more evident than in the honeymoon period following a president s election. The midterm is not a presidential honeymoon, but neither is it the heat of a presidential election. The approval question means different things in different contexts and it should be interpreted in this light. 4 President Obama s approval rating in 2014 was two points lower than it was at the 2010 midterm (45%) in which Democrats lost 64 House seats. Two other modern presidents had similar or identical midterm approval ratings to Obama s in In 1950, President Truman s approval rating stood at 41% and Democrats lost 28 seats. In 1982, President Reagan s approval rating was 43% and Republicans lost 27 seats (Gallup Poll 2014a).

7 The Republican Wave of to advise Republicans that they need to have a serious talk with themselves, and they need to change. Dionne was not alone in this perspective. Ruy Teixeira and John Halpin asserted that the 2012 election marks the culmination of a decades-long project to build an electorally viable and ideologically coherent progressive coalition in national politics (Teixeira and Halpin 2012). As they saw it, the election was a turning point: the age of Reagan and extreme conservatism has given way to the age of Obama and pragmatic progressivism. Still others offered similar takes, often anchoring them in the cement of socio-demographic changes that threatened whether Republicans could remain nationally competitive (Mayer 2014, p. 104 and 110). Others interpreted the 2012 election as turning on the job performance of the president. From this perspective, President Obama was reelected because a majority of the electorate was satisfied with the state of the economy. John Sides and Lynn Vavreck, for instance, claimed that the objective economy favored Obama (2013, pp ), that little during the election year pushed the outcome far from what the economy would predict (2013, p. 224), and that Obama s victory was consistent with the underlying economic fundamentals (2013, p. 236). Eric Fehrnstrom, the advisor to Mitt Rommey who is perhaps forever to be remembered for his unfortunate Etch-a-Sketch metaphor, goes almost as far as Sides and Vavreck in tracing Obama s reelection to the economy. In his comments during Annenberg s post-election debriefing of campaign insiders, Fehrenstom suggests that President Obama won the election because the economy was improving. By his reading, the economy got better just enough for the President to make a persuasive case for staying the course (Jamieson 2013, p. 71, 79). Neither of these explanations is supported by the evidence. The outcome of the 2012 election was not about ideology nor was it about the economy. I discussed the evidence about these interpretations previously (Campbell, 2013a,b), but they bear revisiting in brief. The 2012 election was not an endorsement of liberalism, much less the ushering in of a progressive era. Far from it. Self-described conservatives outnumbered self-described liberals in the 2012 exit polls by ten percentage points (35% to 25%). Those who thought government was doing too much similarly outnumbered those who thought that the government was doing too little (51% to 43%). Even among those who said that they had voted for President Obama, non-liberals outnumbered self-described liberals in the exit polls. There is not a whiff of evidence that the 2012 election marked the triumph of liberalism. President Obama s reelection was also not a positive review of his economic record. Tables 1 and 2 make long-run and short-run comparisons of economic growth during President Obama s first term and during the administrations of

8 616 James E. Campbell Table 1 Economic Growth Over a President s Term for Presidents Standing for Another Term, Rank President and Re-election Year GDP Growth over Term (%) Election Outcome 1. Kennedy/Johnson, Won 2. Ronald Reagan, Won 3. Dwight Eisenhower, Won 3. Bill Clinton, Won 5. Richard Nixon, Won 6. George W. Bush, Won 7. Jimmy Carter, Lost 8. Barack Obama, Won 9. George H.W. Bush, Lost 10. Nixon/Ford, Lost The Real GDP Growth Rate is the mean of quarterly growth rates from the first quarter of a president s second year in office through to the third quarter of the fourth year. The data are from the Bureau of Economic Analysis (2014). other modern presidents who sought reelection. Table 1 takes a long-run perspective, ranking presidents since Eisenhower by the economic growth over their term. This excludes the first year of the president s term as a start-up year with economic conditions inherited from the preceding president. The growth over the term is calculated as the mean real GDP growth (quarterly data) from the first quarter of a president s term through to the third quarter of the reelection year. Table 2 takes the short-run perspective, ranking presidents since Truman by real GDP growth in the second quarter of the reelection year. Election forecasters have found that second-quarter election year growth rates are the strongest economic precursors of the vote. The take-away point from both of these economic rankings is that every modern president going into an election with a weaker economy than President Obama s in 2012 ended up losing his election, and some presidents with even stronger economies than President Obama s lost their elections. 5 In short, the historical record demonstrates quite clearly that no modern president had been elected with an economic record as weak as President Obama s. Annual real GDP growth in 2010, 2011, and 2012 never exceeded an anemic 2.5%. The real GDP growth rates for the four quarters of 2012 were 2.3%, 1.6%, 2.5%, and 0.1%. In a 5 Either adding growth from the third quarter of the election year to the second quarter or adding both the first and third quarters to the first quarter moves President Obama s ranking among the eleven most recent presidents up from tenth place to ninth place. In short, by any historical comparison, economic growth during President Obama s first term was comparatively quite weak.

9 The Republican Wave of Table 2 Economic Growth in the Second Quarter of Reelection Years, Rank President and Re-election Year GDP Growth in 2nd Quarter (%) Election Outcome 1. Richard Nixon, Won 2. Ronald Reagan, Won 2. Bill Clinton, Won 4. Harry Truman, Won 5. Lyndon Johnson, Won 6. George H.W. Bush, Lost 7. Dwight Eisenhower, Won 8. Gerald Ford, Lost 9. George W. Bush, Won 10. Barack Obama, Won 11. Jimmy Carter, Lost The GDP data are from the Bureau of Economic Analysis (2014). word, dismal. Michael Lewis-Beck and Charles Tien characterized the pre-election economic conditions as lackluster economic growth (2012, p. 625). Far from being a political asset, the economy was a liability, a liability so great that economies this weak had consistently led to incumbent defeats up to Exit polls and other pre-election surveys also indicated that voters perceived the economy as weak and not appreciably improving. The exit polls in 2012 indicated that 55% of voters thought that the economy was poor and not improving (25%) or was actually getting worse (30%). This is not a record any incumbent would want to run on. As Robert Erikson concluded in his post-election review, without a doubt, the economy hurt Obama (2014, p. 16). The outcome of the 2012 presidential election depended not on ideology and not on the economy, but on incumbency. Presidents who follow a president of the opposing party (first party-term presidents) have many advantages, and, because of these advantages, they are rarely defeated (Campbell 2008, 2013a,b; Abramowitz 2013; Erikson 2014). Since 1900, there have been twelve first party-term presidents who sought reelection. Only one of the twelve was defeated (Carter in 1980). One of the chief advantages of first party-term presidents is that their predecessors can be plausibly blamed for persisting problems. This was President Obama s salvation. President Obama was reelected despite a poor economy because voters did not blame him for it. They blamed his predecessor, President Bush. The economy in 2012 was in poor shape (though not as catastrophically poor as President Carter s economy). Voters understood this, but they blamed President Bush and gave President Obama more time to get the economy back on track. The 2012 election was not a vote of confidence, not a positive referendum. It was a postponed or suspended referendum.

10 618 James E. Campbell Changes in the Electorates? With the critical backdrop of the 2012 election reviewed, we can return to the puzzle: what changed between then and the 2014 midterm that would create the Republican wave? One possible explanation of the 2014 midterm outcome is that turnout made the difference, that the demographics of the smaller midterm electorate compared to the demographics of the larger presidential electorate favored Republicans in the midterm and Democrats in the presidential year. Since Democrats draw more votes from young voters, racial and ethnic minorities, and women, Republicans would naturally fare better with the older white male midterm electorate. There are two issues within this proposed explanation. The first is whether the two electorates differed substantially, and the second is, to the extent that there were differences, what was the reason for these? Were they based on features of each party s coalition or did they reflect changing political climates? Table 3 presents some exit poll comparisons of the 2012 and 2014 electorates. What is most striking about the two electorates is there similarity to each other. The 2014 electorate was slightly more male, slightly whiter, and slightly more educated. In most cases, the differences were only two or three percentage points. This similarity is particularly notable because of the unusually low turnout in 2014 and, therefore, the unusually big difference in size between the two electorates. The only characteristic on which two electorates differed substantially was age. The 2014 midterm electorate was a good deal older. Though this might reflect President Obama s personal appeal to younger voters in 2012 or their disappointment with the Democrats going into 2014, younger voters have historically been a good deal less likely to vote in midterm elections. In examining exit polls from 1976 to 2010, Best and Krueger found that the proportion of 18- to 29-year olds in the exit polls has dropped roughly 6.5 percentage points on average in midterm election years (2012, p. 55). Comparisons of the median age of presidential and midterm electorates from 1956 to 1994 also found midterm electorates to be routinely older (Campbell 1997). However, even if the age difference has been routinely observed across election cycles, it might be a midterm disadvantage for Democrats, since younger voters have been particularly inclined to vote for Democrats in recent elections. Although the sizeable age gap raises the possibility of a demographic midterm disadvantage for Democrats, the ideological similarities of the 2012 and 2014 electorates suggests that this demographic difference did not make much of a political difference. Whether examining self-reported ideology or views about whether government was doing too little or too much, voters in the 2014 midterm

11 The Republican Wave of Table 3 Demographics and Attitudes of the 2012 and 2014 Electorates (Percentages). Exit Poll Response 2012 Presidential Election 2014 Midterm Election Gender Female Male Age 44 and Under Over Race White Non-White Education College Graduate No College Degree Political Ideology Conservative Moderate Liberal Scope of Government Attitude Should Do More Does Too Many Things Source: The 2012 Exit Poll data are from CNN (2012). The 2014 Exit Poll data are from ABC News (2014). election were slightly more conservative than in the 2012 on-year electorate but only slightly so. Still, the demographic and ideological differences are in the direction that would favor the Republicans and what might appear to be small aggregate differences can make a big difference in outcomes. This raises the second issue about compositional differences. To what extent do they simply reflect different political climates of the two elections? Turnout decisions are influenced by reactions to the political choices and climate of an election as well as by the individual characteristics and interests of potential voters. For example, a young voter with marginal political interest and a Democratic predisposition may have been enthusiastic enough to vote for President Obama and congressional Democrats in the 2012 election, but disgruntled or not inspired enough to turn out in Conversely, a Republican-inclined voter may have been put off by Mitt Romney to the point of not voting in 2012, but then activated to vote in 2014 by opposition to the policies and record of the Obama presidency. This partisan turnout effect is the driving force behind the cycle of presidential surge and midterm decline (Campbell 1987, 1997). The political climate in a presidential year draws out partisans of the winning presidential party and

12 620 James E. Campbell depresses turnout in the losing presidential party. The political climate in the midterm, though weaker, has a similar effect. Turnout in the party favored by the political climate is likely to be higher than it otherwise would be and turnout in the party in disfavor is likely to suffer a bit. In the cycle, the candidates and issues created a more Democratic electorate in the presidential year and a less Democratic electorate in the midterm. Democrats are not always favored in on-years and Republicans are not always favored in midterms. The partisan compositional difference simply reflected the political climates of the two elections, particularly the presidential election. The demographic differences observed in Table 3 are perfectly consistent with the revised theory of surge and decline, the repercussion effects of the presidential election and the midterm referendum on the presidential party s job performance. It does not, however, answer the question of why the midterm referendum on the president s party was so negative that it generated the Republican wave (with a somewhat more Republican oriented set of voters). What changed between 2012 and 2014? The Buck Eventually Stops Here The most politically important change between 2012 and 2014 was the simple passage of time and with it a change in the public s view of who should be held accountable for the continuingly weak economy. A majority of voters blamed President Bush for poor economic conditions in By 2014, 6 years into his presidency and 6 years after his predecessor had left office, responsibility for the economy fell to President Obama and his party. The economy was by no means the only issue on voters minds when they went to the polls in Among the other issues of concern to voters were the impasse over illegal immigration, the future of Obamacare, the threat of terrorism, and continuing conflicts in the Middle East. Although the economy was not the only concern of voters, as in 2012 and as in most elections, it continued to be their greatest concern (Polling Report 2014a). A late September Gallup Poll found that 88% of registered voters in 2014 thought that the economy as an issue in the election was extremely important or very important (Gallup News Service 2014). Although there were some glimmers of improvement with strong second- and third-quarter real GDP growth rates prior to the election (4.6% and 3.9% respectively), declining unemployment rates (down 2 points from 7.8% to 5.8% from the 2012 election), and indications of slightly improving consumer confidence, it was all too little and too late. Combined with a few festering chronic problems (immigration and the Middle East situation), the electorate s patience had worn

13 The Republican Wave of thin. Table 4 reports the economic comparisons between 2012 and Though finishing stronger near the end of 2014, overall economic growth in the three pre-election quarters of the two election years was only about two percentage points. This is a full percentage point weaker than the median economic growth since Recalling Tables 1 and 2, the economies in 2012 and 2014 also compare unfavorably to economic growth rates of presidents who had gone on to win reelections. One could even look to the long-running open-spigot, low-interest monetary policy of the Federal Reserve Board as official acknowledgment that the economy in 2014 was still quite fragile. The economy in 2014, like that in 2012, was struggling. Despite some bits of good economic news, the sluggish condition of the economy was reflected in the electorate s general perceptions. Whether in Gallup Polls before the election or exit polls on Election Day, the public judged the economy to be generally weak in both years. An NBC News and Wall Street Journal survey conducted in early August of 2014 reported that nearly half of Americans (49%) still considered the economy to be in recession (Schrager 2014). After weathering a protracted anemic Table 4 Retrospective Conditions and Evaluations in 2012 and 2014 (Percentages). Conditions, Perceptions, and Evaluations 2012 Presidential Election 2014 Midterm Election Real GDP Growth in Election Year Mean for Quarters 1 3 of Election Year Difference from Median Growth (+3.2%) Perceptions of the Economy (Gallup) Excellent/Good Poor Perceptions of the Economy (Exit Poll) Getting Better or Good and Staying the Same Poor and Staying the Same or Getting Worse Presidential Approval of Handling of the Economy Approve Disapprove Presidential Approval (General) Approve Disapprove Sources: The real GDP growth rates are calculated by the author from Bureau of Economic Analysis (2014). The perceptions of the economy data are from Gallup (2014b). The exit poll data are from CNN (2012) and ABC News (2014). The Presidential Approval Ratings for Handling of the Economy are the mean of Gallup, NBC and Wall Street Journal, CBS News, and ABC and Washington Post polls (Polling Report 2014b). The median real GDP growth (3.2%) is the median of all quarters from the second quarter of 1947 to the third quarter of 2014.

14 622 James E. Campbell recovery for 6 years, Americans were not convinced that the economy had turned the corner. The same NBC News and Wall Street Journal survey reported that 22% of Americans thought the nation was still experiencing difficult economic times and things have not improved that much and another 52% said that the economy has improved a bit, but we still have a long way to go before things get better. Nearly two thirds of respondents (64%) in the survey indicated that they were somewhat or very dissatisfied the state of the economy (Hart Research Associates and Public Opinion Strategies 2014). The key difference between 2012 and 2014 was not the comparative condition of their economies; both were weak. The key difference was whether the incumbent president and his party should be held accountable for the weak economy. In 2012, President Obama had not been in office quite 4 years, and his predecessor President Bush could plausibly be blamed for the poor economic conditions. In the exit polls of 2012, 53% of voters placed more of the blame for the poor economy on President Bush. Only 38% considered President Obama largely responsible. President Obama said at the time that he would give himself an incomplete for his handling of the economy. At the Democratic National Convention that year, former President Clinton made the case that no president could have turned the economy around from the Great Recession in just one term. More time was needed. Voters listened and gave the president and his party more time. They were not happy, but they were patient. This was reflected in President Obama s economic and overall approval numbers in Two years later, things had not improved much or, at least, not enough, and there was now no real question of responsibility. Nearly 6 years into his presidency, almost 2 years into his second term, political responsibility for the economy resided clearly with President Obama and his party. Voters were still disappointed. Many stayed home. A majority of those who turned out indicated that they did not think President Obama had done a good job, in general, or with the economy, in particular. The referendum on the incumbent s performance had been suspended in 2012, but it was not an indefinite suspension. The buck that did not stop on the president s desk in 2012 found its way there in Although the president was spared the voters wrath, his party could not avoid it. The Lessons of 2014 The 2014 midterm election offers several lessons. First, it reaffirms the revised theory of surge and decline and the midterm referendum in explaining electoral

15 The Republican Wave of change in midterm elections. There was a Republican wave in 2014, but it was exhibited in proportionately smaller changes in the House since prior short-term forces there in 2012 had been muted. The wave produced proportionately larger changes in the Senate since prior short-term forces there in 2008 had been larger. In a broader but related sense, the 2014 experience also indicates that elections can only be understood in the context of the preceding election. In the case of 2014, it could partly be understood by its change from the 2012 presidential election (the decline from the surge), but also by its continuity with that prior election (the suspended referendum). The second lesson of 2014 may be that election analysts and politicians should be more careful in interpreting elections. Election outcomes are easily misinterpreted, and these misinterpretations may have important political and policy consequences. The individual vote is a blunt enough instrument. Multiply that by millions and you may have a complicated puzzle to figure out. Add intensely felt political motives and enthusiasms by the interpreters plus limited and imperfect data, and you have a real puzzle on your hands. On top of this, add journalistic deadlines that create a rush to bad judgment, and the puzzle dissolves into one big polemical mess unless we step back and carefully evaluate all the evidence we have. Why is this important? The reasons why politicians are elected by voters should affect how those politicians govern. This is not to assume a democratic impulse in politicians, just a self-preservation or reelection motive. If an election is misunderstood as an ideological endorsement ushering in a new era, political leaders may wrongly feel emboldened and supported in pushing a more ideologically ambitious agenda. The result is overreach, bad public policy, and policy that will later be repudiated in following elections. If an election is misunderstood as a positive referendum when it was really a suspended referendum, political leaders may wrongly feel they have done a good job, when the voters say otherwise. The third lesson of 2014 concerns political polarization. As much as observers of American politics have focused in the polarization of the parties and the public, and as important as polarization is to understanding modern American politics, the outcomes of national elections are more often decided by the performance of the parties in handling the valence issues that all Americans care about. Values decide more votes, but performance decides more elections. The reason for this is that there is a rough parity between the parties based on their ideological moorings. There are more conservatives than liberals, but the Republican Party is further to the right than Democrats are to the left. This establishes a very competitive party system. But as polarized as Americans have become, there are still plenty who have only marginal political interests. These are the liberals who

16 624 James E. Campbell turn off of politics when Democrats hit a rough patch and the conservatives who tune out when things are not going so well for Republicans. It is the performance of the parties that shifts voting and turnout of not only swing voters, but less centrist voters as well. The final lesson of the 2014 midterm election concerns first party-term presidents. The 2014 midterm election reaffirms the conclusion that the first party-term incumbency advantage was critical in deciding the 2012 presidential election. If the 2012 presidential election had been a straight-up referendum on the national conditions presided over by the in-party, the 2014 midterm election would have been President Romney s first midterm. Instead, voters held their judgment in abeyance. They gave President Obama more time to get the economy in shape. The lesson of 2014 would seem to be that voters will give a new president a second term to deal with an inherited mess, but they do expect to see clear results by a president s second midterm election. References ABC News Midterm Election Results. Accessed, November 6, Abramowitz, Alan I Voting in a Time of Polarization: Why Obama Won and What It Means. In Barack Obama and the New America: The 2012 Election and the Changing Face of Politics, edited by Larry J. Sabato, Lanham, MD: Rowman and Littlefield. Abramowitz, Alan I Forecasting the 2014 Midterm Elections with the Generic Ballot Model. PS: Political Science & Politics 47 (4): Bean, Louis H The Mid-Term Battle. Washington, D.C.: Cantillion Books. Best, Samuel J., and Brian S. Krueger Exit Polls: Surveying the American Electorate, Washington, D.C.: CQ Press. Bureau of Economic Analysis Gross Domestic Product (GDP): Percent Change from Preceding Period. Accessed October 30, htm=gdp. Campbell, Angus Surge and Decline: A Study of Electoral Change. Public Opinion Quarterly 24 (3)(Autumn): Campbell, James E Explaining Presidential Losses in Midterm Congressional Elections. The Journal of Politics 47 (1): Campbell, James E The Revised Theory of Surge and Decline. American Journal of Political Science 31 (4): Campbell, James E The Presidential Pulse of Congressional Elections. 2nd ed. Lexington, KY: The University Press of Kentucky. Campbell, James E An Exceptional Election: Performance, Values, and Crisis in the 2008 Presidential Election. The Forum: A Journal of Applied Research in Contemporary Politics, 6 (4): article

17 The Republican Wave of Campbell, James E. 2013a. The Miserable Presidential Election of 2012: A First Party-Term Incumbent Survives. The Forum: A Journal of Applied Research in Contemporary Politics 10 (4): Campbell, James E. 2013b. A First Party-Term Incumbent Survives: The Fundamentals of In Barack Obama and the New America: The 2012 Election and the Changing Face of Politics, edited by Larry J. Sabato, Lanham, MD: Rowman and Littlefield. Campbell, James E. 2014a. Introduction: The 2014 Midterm Election Forecasts. PS: Political Science & Politics 45 (4): Campbell, James E. 2014b. The Seats-in-Trouble Forecast of the 2014 Midterm Congressional Elections. PS: Political Science & Politics 47 (4): CNN Exit Polls. Accessed December 10, results/race/president. Dionne, E. J Obama s Victory Should Settle a Bitter Argument. Real Clear Politics, November 8. Accessed November 16, Erikson, Robert S Forecasting the 2012 U.S. Presidential Election: Should We Have Known All Along That Obama Would Win? In The 2012 Presidential Election: Forecasts, Outcomes, and Consequences, edited by Amnon Cavari, Richard J. Powell and Kenneth R. Mayer, Lanham, MD: Rowman and Littlefield. Gallup News Service Midterm Election Fall Benchmark. Accessed December 9, Gallup Poll. 2014a. Presidential Approval Ratings (various presidents). Accessed December 1, Gallup Poll. 2014b. Gallup Daily: U.S. Economic Conditions. Accessed, November 4, Hart Research Associates and Public Opinion Strategies Study = 14643, NBC News/ Wall Street Journal Survey. Jamieson, Kathleen Hall Electing the President 2012: The Insiders View. Philadelphia: University of Pennsylvania Press. Lewis-Beck, Michael S., and Charles Tien Election Forecasting for Turbulent Times. PS: Political Science & Politics 43 (4): Mayer, Kenneth R Lessons of Defeat: Republican Party Responses to the 2012 Presidential Election, In The 2012 Presidential Election: Forecasts, Outcomes, and Consequences, edited by Amnon Cavari, Richard J. Powell and Kenneth R. Mayer, Lanham, MD: Rowman and Littlefield. Oppenheimer, Bruce I., James A. Stimson and Richard W. Waterman Interpreting U.S. Congressional Elections: The Exposure Thesis. Legislative Studies Quarterly 11 (2): Polling Report. 2014a. Problems and Priorities. CBS News Poll of October 23-27, 2014 and CNN/ORC Poll of September 5-7, Accessed December 9, Polling Report. 2014b. President Obama and the Obama Administration. Accessed December 2, Schrager, Allison Nearly Half of Americans Think the Recession Is Not Over. Bloomberg Businessweek. August 19. Accessed December 9, com/ articles/ /the-psychological-damage-of-the-recession-is-not-going-away.

18 626 James E. Campbell Sides, John, and Lynn Vavrek The Gamble: Choice and Chance in the 2012 Presidential Election. Princeton, NJ: Princeton University Press. Teixeira, Ruy, and John Halpin The Return of the Obama Coalition. Center for American Progress, Accessed November 16, progressive-movement/news/2012/11/08/44348/the-return-of-the-obama-coalition/. Tufte, Edward R Determinants of the Outcomes of Midterm Congressional Elections. American Political Science Review 69 (3): James E. Campbell is a UB Distinguished Professor of Political Science at the University at Buffalo, SUNY. He is the author of three books, including The American Campaign: US Presidential Campaigns and the National Vote, and more than 80 journal articles and book chapters. His election analyses have appeared in The Forum since the 2004 presidential election.